Report Interpretation

Overview

Taking Zhuque-3's completion of China's first land-based recovery using landing legs as its starting point, the report compares the recovery architectures, technical performance, and cost structures of Zhuque-3, Long March 10B, and SpaceX. Its central conclusion is that China has crossed the threshold of whether recovery is possible, but has yet to demonstrate that first-stage rockets can be reused frequently, reliably, and economically. Scale benefits still depend on improvements in payload ratio, system iteration, and stable order density.

Core views

Zhuque-3 completed first-stage rocket recovery on August 19, 2026, marking China's first landing-leg recovery and the first land-based recovery of an orbital-class first-stage rocket. According to the company, its reusable liquid oxygen-methane launch vehicle has established a reuse-oriented technological framework spanning overall rocket design, propulsion systems, return control, stainless-steel rocket body manufacturing, and ground support. Nomura's first-level assessment is that China's two recovery approaches will stimulate different supply chains: offshore net-catching will concentrate spending on marine equipment such as recovery vessels, dynamic positioning, and nets, while land-based landing-leg recovery will retain spending on the rocket itself, primarily involving landing mechanisms, cushioning, terminal control, and thermal protection. Zhuque-3's land-based recovery approach is relatively easy to replicate. It relies on onboard landing legs to land in a predetermined terrestrial area, while the ground side only needs to prepare the landing zone and does not require a capital-intensive recovery fleet. Expansion therefore mainly depends on the number of landing sites, and the report expects other private operators to be more likely to adopt this approach. There are two trade-offs: the landing legs themselves add dead weight; and inland launches must avoid densely populated debris fall zones, narrowing available launch azimuths and causing rockets to deviate from energy-optimal trajectories. During recovery, vessels can adjust their positions in real time, whereas fixed land-based landing sites can only expand their clearance zones. Zhuque-3 must fly precisely toward fixed safe coordinates rather than the lowest-energy landing point. Based on industry research, the report judges that this will tighten trajectory design and reduce mission applicability and payload capacity. Long March 10B has chosen the opposite trade-off: replacing landing legs with hooks and a net-catching system on a dedicated vessel, shifting costs to approximately CNY0.8–1.0bn of offshore infrastructure and approximately CNY5mn of consumables per mission, while recovering approximately 1 tonne of payload capacity on a roughly 500-tonne-class rocket. The constraint of this approach is that, after offshore recovery, the rocket must still be transported 400 to 600 kilometers and returned to the workshop over 1 to 2 days, limiting turnaround speed and launch frequency. The risks are also asymmetric: a landing accident on an ordinary drone ship may be tolerable, but an error could destroy a net-catching platform worth approximately CNY1bn. By comparison, Falcon 9 uses landing legs and a fleet of drone ships, has achieved a 97% to 98% recovery success rate, has reused a single booster as many as 33 times, and has completed more than 500 recoveries in total. As of August 2026, China's Long March 10B and Zhuque-3 had completed only two first-stage recoveries combined, and Zhuque-3 had yet to fly again. The report's second-level assessment is that Zhuque-3 can already land in a manner similar to Falcon 9, but does not yet possess Falcon 9-like flight performance. Chinese engines have combustion chamber pressure of approximately 10MPa, versus Merlin's 25MPa and Raptor's more than 30MPa; SpaceX engines have a thrust-to-weight ratio of approximately 240, versus less than 120 for other solutions. Higher combustion chamber pressure and thrust-to-weight ratios mean less engine and structural weight is required for the same thrust. Consequently, Falcon 9 achieves a payload ratio of 4.2%, compared with 1.8% to 3% for other industry solutions. Zhuque-3 currently corresponds to a payload of 5 to 6 tonnes, while the nominal comparison value shown in the report is 21.3 tonnes. The report estimates its current cost at CNY28,000 to 30,000 per kilogram, higher than the price of expendable rockets overseas. The report's market survey indicates that recoverable designs consume approximately 30% of payload capacity. This loss is manageable when starting from a payload ratio of approximately 4%, but achieving viable economics is difficult when starting from approximately 2%. The cost gap is summarized in terms of pricing, technology, and scale. Pricing is the most difficult factor to replicate: the report's survey indicates that the high profitability of SpaceX's launch business may arise from internal costs below USD40mn, while external pricing for military and commercial contracts reaches USD100–200mn. Reliability validation and certification increase mission costs, but the incremental cost is relatively small given SpaceX's high launch frequency. The technological cost advantage does not arise solely from the performance of individual components. The report states that SpaceX uses system-level fault tolerance rather than pursuing extreme reliability in every component, allowing industrial-grade parts and lower-cost materials to replace some aerospace-grade products. At the same time, it uses structural standardization to unify the nearly 20 sections of a rocket and their end-interface frames into specifications suitable for mass production. The root cause of the scale gap is demand and order density. SpaceX completed more than 160 missions in 2025, compared with only 10 to 20 missions annually a decade earlier, enabling a shift from small-batch manufacturing to industrialized production. The report believes vertical integration generally requires more than approximately 100 launches per year to spread costs effectively. A single Chinese rocket model launches only 10 to 15 times annually, while the threshold for economies of scale is more than 30 launches. SpaceX also has an internal order book generated by Starlink and Starshield, whereas Chinese rocket companies must bid for each mission individually and follow customer schedules. Therefore, even if the technological approach can catch up, insufficient demand will continue to constrain manufacturing scale, turnaround rates, and reuse amortization. The first-stage rocket accounts for approximately 65% of Falcon 9's total construction cost, and its cost can fall to a single-digit percentage only after the same booster has flown repeatedly. For Starship, the first and second stages account for approximately 55% and 45% of costs, respectively. The report's main text estimates that an expendable Falcon mission costs approximately USD67mn, while a mission using a booster that has flown more than 30 times costs approximately USD34–37mn, which it states is approximately 40% of an expendable mission's cost. A separate amortization model in the report's chart shows that, based on a total cost of USD67mn, a first-stage cost of USD43.6mn, and other costs of USD23.5mn, reusing the first stage 30 times reduces its per-mission cost to USD5.7mn and total mission cost to USD29.1mn, equivalent to 43.5% of the expendable cost, with the first-stage cost index falling to 13. In terms of timing, SpaceX first recovered a first-stage rocket in 2015 but did not achieve reliable, high-frequency reuse until after 2020. If Zhuque-3 follows the same five-year path beginning in 2026, the maturity milestone would occur around 2031. Nomura's industry survey is slightly more optimistic, reasoning that followers do not need to explore the entire path again: establishing reuse clearance standards may take approximately 1 to 2 years, while achieving viable unit economics and batch reuse and reducing prices to CNY8,000 to 10,000 per kilogram may take approximately 3 to 4 years, corresponding to 2029 to 2031. Therefore, successful recovery is an important starting point, but a high payload ratio, reliable reflights, a sufficient number of reuses, and stable order density are all necessary conditions for a viable reuse business model.

Analysis framework

The report first compares three recovery architectures—land-based landing legs, offshore net-catching, and landing legs combined with drone ships—to identify where capital expenditure, onboard weight, transportation turnaround, and accident tolerance reside in each approach. It then measures the technological gap using combustion chamber pressure, engine thrust-to-weight ratio, payload ratio, and recovery records, before constructing an amortization curve based on the first-stage rocket's share of cost and the number of reuses. Finally, combining industry surveys with SpaceX's historical path, the report explains differences in unit costs through contract pricing, manufacturing methods, launch frequency, vertical integration, and internal order density, and uses this analysis to estimate when China may achieve economically viable batch reuse.

Methodology notes

• Industry/Sector Analysis FrameworkUpstream-Midstream-Downstream Supply Chain Transmission

  Recovery approach and supply-chain load analysis

  The report compares land-based landing legs with offshore net-catching and maps the expenditure associated with each technological approach either to onboard mechanisms, control, and thermal protection or to marine equipment such as recovery vessels, positioning systems, and nets, in order to assess the impact of different approaches on each segment of the supply chain.

• Industry/Sector Analysis FrameworkCost curve analysis

  First-stage rocket reuse amortization curve

  The report amortizes first-stage rocket costs over the number of reuses and adds costs associated with each reflight to compare per-mission costs between expendable launches and different reuse scenarios.

• Competition and Strategy FrameworkEconomies of scale / learning curve

  Scale threshold for launch frequency and order density

  Using SpaceX's increase from 10 to 20 missions annually to more than 160 missions in 2025, the report demonstrates that mass manufacturing, standardization, and vertical integration can spread costs only when supported by sufficiently high launch frequency and stable orders.

• Industry/Sector Analysis FrameworkVolume-price decomposition

  Decomposition of payload capacity and launch cost per kilogram

  The report considers payload tonnage, payload ratio, per-mission cost, and per-kilogram pricing together to explain how the loss of carrying capacity caused by recovery mechanisms affects final unit economics.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Zhuque-3 (LandSpace, unlisted)

  The core example of China's land-based landing-leg recovery approach, which completed first-stage rocket recovery on August 19, 2026.

  Strengths

  Land-based recovery does not require a dedicated fleet, with the ground side primarily needing to prepare a landing area, resulting in lower capital expenditure and making the approach easier for other private operators to replicate.

  Weaknesses

  Landing legs add weight, while fixed landing zones and inland safety requirements constrain launch azimuths, trajectory design, mission scope, and payload. It has completed only one recovery and has yet to fly again.

  Comparison

  It can already land in a manner similar to Falcon 9, but still has significant gaps in engine pressure, thrust-to-weight ratio, payload ratio, number of recoveries, and launch frequency.

  Risks

  If successful recovery cannot be converted into reliable reflights, a higher payload ratio, and sufficient order density, unit costs may remain uncompetitive for an extended period.

• Long March 10B

  Represents an alternative Chinese recovery approach that uses hooks and a dedicated vessel-based net-catching system for the first-stage rocket.

  Strengths

  Eliminating landing legs can recover approximately 1 tonne of payload capacity on a roughly 500-tonne-class rocket.

  Weaknesses

  It requires CNY0.8–1.0bn of offshore infrastructure, approximately CNY5mn of consumables per mission, and an additional 400 to 600 kilometers and 1 to 2 days of transportation.

  Comparison

  Compared with Zhuque-3, Long March 10B shifts recovery costs from onboard mechanisms to offshore infrastructure, exchanging higher capital expenditure for payload benefits.

  Risks

  A landing error could destroy a net-catching platform worth approximately CNY1bn, while maritime transportation turnaround constrains launch frequency.

• SpaceX (SPCXUS, Not rated)

  Serves as the primary benchmark for China's reusable rocket technology, costs, and economies of scale.

  Strengths

  Falcon 9 has achieved a recovery success rate of 97% to 98%, with a single booster reused up to 33 times and more than 500 cumulative recoveries. Mission volume exceeded 160 in 2025, supported by internal orders from Starlink and Starshield.

  Weaknesses

  The report believes its high profitability partly depends on external pricing of USD100–200mn for military and commercial contracts, a condition that is difficult for Chinese operators to replicate.

  Comparison

  Its engine combustion chamber pressure, thrust-to-weight ratio, and 4.2% payload ratio are all clearly superior, while its scale and order density also exceed the 10 to 15 annual launches of a single Chinese model.

Key data

• Zhuque-3's first land-based recoveryAugust 19, 2026China's first landing-leg recovery and the first land-based recovery of an orbital-class first-stage rocket
• Long March 10B offshore net-catching infrastructureCNY0.8–1.0bnCapital expenditure for dedicated vessels and net-catching systems, plus approximately CNY5mn of consumables per mission
• Payload benefit of the net-catching approachApproximately 1 tonnePayload capacity recovered by eliminating landing legs on a roughly 500-tonne-class rocket
• Offshore recovery transportation constraints400–600 kilometers, 1–2 daysAdditional transportation required to return the first-stage rocket from the offshore recovery area to the workshop
• Engine combustion chamber pressureChina approximately 10MPa; Merlin 25MPa; Raptor more than 30MPaUsed by the report to explain differences in engine and structural weight
• Engine thrust-to-weight ratioSpaceX approximately 240; other solutions below 120A higher thrust-to-weight ratio helps improve the payload ratio
• Payload ratioSpaceX 4.2%; other solutions 1.8%–3%The report believes this metric determines whether the loss of carrying capacity caused by reuse can be tolerated
• Falcon 9 recovery performance97%–98% success rate; up to 33 reuses; more than 500 cumulative recoveriesCompared with Zhuque-3, which has completed only one recovery and has yet to fly again
• SpaceX launch business costs and pricingInternal costs below USD40mn; external pricing of USD100–200mnPricing range for military and commercial contracts cited in the report
• Launch frequencySpaceX completed more than 160 launches in 2025; a single Chinese model launches 10–15 times annuallyThe report's survey indicates that the threshold for economies of scale is more than 30 launches per year, while vertical integration requires approximately more than 100
• Falcon expendable mission costApproximately USD67mnThe first-stage rocket accounts for approximately 65% of total rocket construction costs
• Falcon high-reuse mission costMain-text estimate of USD34–37mn; USD29.1mn in the 30-use amortization model scenarioThe model's 30-use scenario is equivalent to 43.5% of the expendable cost
• Zhuque-3's current unit costCNY28,000–30,000 per kilogramThe report states that this is higher than the price of expendable rockets overseas
• Target period for China's batch-reuse costs2029–2031E, CNY8,000–10,000 per kilogramThe industry survey expects 1 to 2 years to establish reuse clearance standards and 3 to 4 years to achieve viable unit economics

Impact & implications

The report believes Zhuque-3's successful recovery demonstrates the engineering feasibility of China's land-based landing-leg recovery approach and may encourage private operators to direct more supply-chain spending toward onboard landing mechanisms, cushioning, terminal control, and thermal protection rather than capital-intensive offshore platforms. However, industry value will not be realized solely through the first recovery: engine and structural performance determine the payload ratio; reflight reliability and the number of reuses determine whether first-stage rocket costs can be sufficiently amortized; and order density determines whether standardized production and vertical integration have a sufficient scale foundation. The report therefore views 2029 to 2031 as a window in which viable unit economics and batch reuse may gradually emerge, rather than equating the 2026 recovery event with commercial maturity.

Risks

• Fixed land-based landing zones narrow launch azimuths and force rockets to deviate from energy-optimal trajectories, thereby constraining mission scope and payload.
• The offshore net-catching approach requires high capital expenditure and long transportation cycles, while a landing error could destroy a high-value recovery platform.
• If Chinese engine performance and payload ratios improve slowly, the loss of carrying capacity caused by recovery may make viable unit economics difficult to achieve.
• The launch frequency and order density of a single Chinese model are below the threshold for economies of scale, potentially delaying standardized production, vertical integration, and reuse amortization.

What to watch

• Whether Zhuque-3 can progress from a single successful recovery to reliable reflights and batch reuse.
• Whether clear reuse clearance standards can be established over the next 1 to 2 years.
• Whether engine combustion chamber pressure, thrust-to-weight ratio, and overall rocket payload ratio can continue to improve.
• Whether the annual launch frequency of a single Chinese model can exceed the report's stated threshold of more than 30 launches for economies of scale.
• Whether launch costs per kilogram can fall to CNY8,000 to 10,000 in 2029 to 2031.